[R] AD-model builder under R for mix logistics model, a problem.
the code is list below , i can run this code for laplace approximation
but not importance sampling. the code work properly if use impSamp=0 ,
but error for any other number >0 .
could you guys help me out.
thanks so much !
source("http://www.mi.uib.no/~skaug/cash/lesaffre_dat.s";)
library(glmmADMB)
fit_glmmADMB_ip <- glmm.admb(y~treat,random=~1,group="subject",impSamp
=100,data=lesaffre,family="binomial",link="logit",save.dir= NULL)
Initial statistics: 2 variables; iteration 0; function evaluation 0
Function value 1.5926934e+003; maximum gradient component mag 1.2500e+001
Var ValueGradient |Var ValueGradient |Var Value
Gradient
1 0. 1.2500e+001 | 2 0. 6.2441e-001 |
- final statistics:
2 variables; iteration 9; function evaluation 13
Function value 1.2593e+003; maximum gradient component mag -5.7984e-005
Exit code = 1; converg criter 1.e-004
Var ValueGradient |Var ValueGradient |Var Value
Gradient
1-56.9607 -1.2327e-005 | 2 -3.7159 -5.7984e-005 |
inner maxg = 9.681e-005
max separable g 9.681e-005
Newton raphson 1
f = 971.7651636 max g = 9.245773036e-010
max separable g 9.245775812e-010
Newton raphson 2
f = 971.7651636 max g = 2.220446049e-015
971.7651636
Error in glmm.admb(y ~ treat, random = ~1, group = "subject", impSamp =
100, :
The function maximizer failed
In addition: Warning message:
In shell(paste(.path.package("glmmADMB"), "/admb/", file_name, ".exe", :
'cmd' execution failed with error code 1
>
__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
[R] AD-MB problem in package:glmmADMB
Hi here,
i recently notice the software built by otter company know as AD-MB
tools . i explore a little bit this tools under the R Package for mixed
logistics model. it is a very interesting tool. but i get several questions.
first question: i tried the R code below, and have the error , i can run
the code for switching the setting to impSamp=0 , it means i can only
run the code under Laplace approximation but importance sampling. does
anyone have same problem here ?
second question: although i used save.dir=NULL option , but i can not
see the folder which saves the intermedia output, i didn't find the
nbmm.std file which save the BLUP. the strangest thing is i can see a
temporary folder name as _glmm_ADMB_temp_dir_ when programing is running
, but this fold mysteriously disappeared after the programing done .
searched the whole computer desn't find any clue.
I posted my output below.
I really appreciate if you can help me with the questions.
source("http://www.mi.uib.no/~skaug/cash/lesaffre_dat.s
")
fit_glmmADMB_ip <- glmm.admb(y~treat,random=~1,group="subject",impSamp
=100,data=lesaffre,family="binomial",link="logit",save.dir= NULL)
Initial statistics: 2 variables; iteration 0; function evaluation 0
Function value 1.5926934e+003; maximum gradient component mag 1.2500e+001
Var Value Gradient |Var Value Gradient |Var Value Gradient
1 0. 1.2500e+001 | 2 0. 6.2441e-001 |
- final statistics:
2 variables; iteration 9; function evaluation 13
Function value 1.2593e+003; maximum gradient component mag -5.7984e-005
Exit code = 1; converg criter 1.e-004
Var Value Gradient |Var Value Gradient |Var Value Gradient
1-56.9607 -1.2327e-005 | 2 -3.7159 -5.7984e-005 |
inner maxg = 9.681e-005
max separable g 9.681e-005
Newton raphson 1
f = 971.7651636 max g = 9.245773036e-010
max separable g 9.245775812e-010
Newton raphson 2
f = 971.7651636 max g = 2.220446049e-015
971.7651636
Error in glmm.admb(y ~ treat, random = ~1, group = "subject", impSamp =
100, :
The function maximizer failed
In addition: Warning message:
In shell(paste(.path.package("glmmADMB"), "/admb/", file_name, ".exe", :
'cmd' execution failed with error code 1
__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
[R] question for Logic Regression
Hi All: how to get the coefficient for logic regression using selection=2 ( fit multiple models) and type=3 ( logistic regression) for example i have a fit like below : fit<-logreg(resp = model.dat[,21], bin=model.dat[, 2:18],sep=model.dat[,1] ,type=3,select=2,ntrees=2,nleaves=6 ,anneal.control=Annealing,tree.control=TreeControl) i try to use fit$coef but i get nothing. and i try to use eval.logreg to evaluate a validate data "model.dat2", but i cannot fit "model" class make below formula work alltrees <- eval.logreg(fit$model , model.dat2) could anyone enlighten me a little. thanks leo __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] some question about partial prediction in survival
Hi there: i got a problem to get the prediction from a model recently. for example if i use a survival analysis to predict the risk. i use the code like below: i found the the prediction is not equal to (coef * x + coef * sex) , could someone help me with why this happened? and can someone explain to me how this command "predict(f, type="terms")" works? is every partial prediction equal to coef*x=predict(f,type="terms")[,1] and coef*sex=predict(f,type="terms")[,2]? it looks like they did not match. however "predict(f)=predict(f,type="terms")[,1]+predict(f,type="terms")[,2]" how should i understand this. thanks so much for your help test1 <- list(time= c(4, 1,1,2,2,3), status=c(1,1,0,1,1,0), x= c(0,1,1,1,0,0), sex= c(0,0,0,1,1,1)) f<-coxph( Surv(time, status) ~ x + sex, test1) Call: coxph(formula = Surv(time, status) ~ x + sex, data = test1) coef exp(coef) se(coef) zp x1.713 5.546 1.34 1.282 0.20 sex -0.154 0.857 1.45 -0.106 0.92 Likelihood ratio test=1.85 on 2 df, p=0.397 n= 6 bests; T.D __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
